Docstoc

Virtual Humans (PowerPoint)

Document Sample
Virtual Humans (PowerPoint) Powered By Docstoc
					Agents Animations Avatars
AKI-minor 2001-2002

Virtual Humans & Computer Science
• Graphics – algorithms • Human Computer Interaction – interfaces • Artificial Intelligence – rational agents

Overzicht
1 2 3 4 5 Kunstmatige Intelligentie Clockwork bodies Cognitive Science Modellen van Emoties Digitale lichamen: Agents Animations Avatars

Artificial Intelligence

The study is to proceed on the basis of the conjecture that every aspect of learning or any other feature of intelligence can in principle be so precisely described that a machine can be made to simulate it. Mc Corduck 1979

Aims
• Produce on a computer, a pattern of output that would be considered intelligent if displayed by human beings. • Program: test of a particular theory of how a cognitive process might work. • Strong AI: the appropriately programmed computer is a mind.

2
Clockwork bodies

Mechanics
• René Descartes (1596-1650):
– interested in automata that could simulate the human body – non-human animals are like machines

• Julian Offray de la Mettrie (1747):
– the brain is subject to study and duplication

• Jacques de Vaucanson:
– mechanical flute players, ducks

Wim Delvoy Cloaca foto: Dirk Pauwels

3
Cognitive Science

Mindware
• • • • • Leibniz George Boole Charles Babbage Alan Turing John von Neumann • • • • • Calculemus Laws of thought Calculating Engines Turing Machine Computer Architecture

Intelligence
• Software versus Hardware Agents (robots) • Agent: – Agency = active behavior – Rational actions – Autonomy versus Control • Cognition: beliefs - desires - intentions (plans) • Perception - Interpretation - Action • Emotional Intelligence

Intelligence = Problem solving
• • • • • • • Chess Measured as with IQ tests Solve Logic puzzles Learning Expert Systems: medical diagnosis Knowledge of the world: Cyc Drawing conclusions

Turing test (1950)
I propose to consider the question “Can machines think?” This should begin with definitions of the terms “machine” and “think.” Instead of attempting such a definition I shall replace the question by another, which is closely related to it and is expressed in relatively unambiguous words. The new form of the problem can be described in terms of a game which we call the “imitation game.”

Embodied Conversational Agent
• Agent – autonomous – rational – software • Conversational – Natural Language Dialogues • Embodied – Non-Verbal Communication

Face-to-face Conversations
• • • • • Highly contextualized form of language use Employing several channels simultaneously Type of joint action by multiple individuals Complex process of managing interactions Structured by a number of conventions or rules

Gaze Behavior

Gaze Detection

Interaction in VR: Agents

4
Modellen van Emoties

5
Digital Bodies

GOTO WELBO

JACOB: Teaching Agent
• Teaches “Towers of Hanoi”
• Monitors user, corrects and interacts about progress • Demonstrates when asked

ANANOVA
• Virtual newscaster • 28 years old, 5 ft 8 ins tall • Passion for news and information • Loves football and cricket (statistics) • Knows she is different and understands her limitations • Fear of electronic bugs • Struggles with concepts like love and hate

Interactions in VR: Agents
Jennifer James, a saleswoman Specific domain knowledge Superficial global knowledge Builds relationships of affection, trust and loyalty with her customers

Lara Croft

Kirsten Geisler

Webbie Tookay

KYOKO DATE
• Walking, talking, virtual girl • Launched in 1996 by HoriPro Inc., Japan, • Made by Visual Science Laboratory • CD single “Love Communication”

Virtual Idols: Kyoko Date

• parents run a sushi bar • likes reading manga • works part-time fast food restaurant

Idea

• The avatars in this version attempt to automate gaze and basic hand motion during a three-way conversation. • The only input is the text typed by the users, no avatar control is needed.

Mechanism
• Touch the keyboard:
– preparation for typing a message – your avatar shows a "wanting-turn" behavior.

• Delivering the message,
– your avatar first looks away from listeners ("planning phase"), – starts speaking and gesturing – and then ends by looking at the next speaker.

• The next speaker is by default the one the user took the turn from.
– If another listener has indicated they want the turn in thE meantime, that user becomes the next speaker. – If the current speaker expicitly names a person that is present, that person is offered the turn.

GOTO BODYCHAT

GOTO SYNTHETIC CHARACTERS

Concluderend
• Kunstmatige Intelligentie • Kunstmensen – Autonomie of Animatie – Agent of Avatar – Interactie Mens-Machine • Modellen en Simulaties • Techniek - Cultuur


				
DOCUMENT INFO